Practice for numpy arrays

[vinu_ta_ds]

try to create the following numpy arrays

1. a 3x3 numpy array that’s like an identity matrix (1 as the element in the diagonal).
2. a 1d array containing just nan values. (length as you prefer)
3. a 2D array with just 1 as values

also, checkout np.where, np.select and np.vectorize. very useful.